基于机器学习的管制员智能排班算法  

Research on Intelligent Scheduling Algorithm for Air Traffic Controllers Based on Machine-learning

作　　者：张勇[1] 岳谭谭[2] 金沙舟 鄢丹青 ZHANG Yong;YUE Tantan;JIN Shazhou;YAN Danqing(The Second Research Institute of CAAC,Chengdu 610041,Sichuan,China;Shandong Airlines CO.,LTD.,Jinan 250014,Shandang,China)

机构地区：[1]中国民用航空局第二研究所,四川成都610041 [2]山东航空股份有限公司,山东济南250014

出　　处：《民航学报》2020年第1期17-20,共4页Journal of Civil Aviation

摘　　要：随着民航事业的高速发展,管制员工作量和负荷也越来越大,为保证管制安全,需制定合理的排班制度以保证管制员有合理的休息时间。本文分析了国内外对于管制员负荷和排班的关系和研究现状,研究了人工排班的缺陷及以往排班系统的局限性,最后提出基于机器学习的管制员排班算法,并利用实例对排班方法进行对比验证,证明方法的可行性和实用性。Workload of air traffic controllers has increased over the years owing to the rapid development of civil aviation. To enhance the safety of air traffic control, a more reasonable scheduling method needs to be put in place to ensure enough rest time for controllers. This paper first reviews previous worldwide studies on linkage between scheduling method and controller workload. It then discusses current scheduling methods and their potential drawbacks. As a potential response to these drawbacks, a machine-learning based scheduling algorithm is proposed, along with a comparative verification to prove its feasibility and practicality.

关 键 词：机器学习 排班方法 管制员 管制运行 

分 类 号：V355.1[航空宇航科学与技术—人机与环境工程]